Download 1 Are monetary policy and fiscal policy gender neutral? Selin Secil

Are monetary policy and fiscal policy gender neutral?
Selin Secil Akin
1. Introduction
In this paper, I focus on the effects of contractionary monetary policy on the ratio of
women’s unemployment rates to men’s for OECD countries. I observe if this policy
have disproportionate impacts on women.
Unemployment is one of the most significant economic problems, and the cause of
some other economic problems. In addition to unemployment, there are also other
problems related to employment such as informal employment, underemployment,
and unpaid domestic labor. Because of these reasons, reducing unemployment was
an important target for macroeconomic policies. As a part of macroeconomic policy,
monetary policy applied by central banks was also responsible from being a part of
employment creation process. Moreover, development policies were also important
for many central banks in the post-Second World War period (Epstein, 2007:2).
However, since 1980s, development or decreasing unemployment is not the
primary objective of economic policies for a lot of countries. Instead reducing
inflation became one of the most important aims of monetary policy. For more than
two decades, the main aim of the central banks of the countries with inflation
targeting became maintaining price stability. The other targets such as development
and employment policies lost their importance in terms of monetary policy.
Women and men employments have different dynamics so unemployment of
women and men should be examined under different groups. Because of patriarchal
norms, males usually have the breadwinner role in the family, whereas females are
1
responsible mostly from reproductive activities, and less engaged in paid work. For
these reproduction activities, women do domestic work, which is unpaid. Labor
force participation rate of women is lower, and unemployment rate of women is
usually higher than men. Even if they participate in labor force, it is usually more
difficult for women to find a job. There are different kinds of explanations to clarify
why women are more unemployed than men1. Moreover, a significant number of
women are employed in informal employment.
Economic policies can be used to increase women employment, and remove the
inequality between women and men in the labor market. If the unemployment
problem is just left to the market conditions, women will continue to be
disadvantages in the labor market. However, if some economic policy tools are used,
positive discrimination can be applied for women, and they can overcome these
disadvantages.
To understand which economic policy can be used to increase women’s employment
rate, firstly we should observe which economic policy causes disproportionate
effects, and which one can reduce these effects. In this respect, as starting point, I
will focus on contractionary monetary policy. This paper aims to look for the
disproportionate impacts of contractionary monetary policy on women employment
for OECD countries for the period 1980-2014, by using unbalanced panel data.
The rest of paper includes four main parts except from concluding remarks. For the
next part, I will discuss the literature review about the impacts of monetary policy
on women’s unemployment rate. Then I will briefly discuss the contractionary
1
For example, Seguino (2004) discusses the causes of women unemployment for Caribbean.
2
monetary policy in general, and the gender differences in the economy in the
following section. Finally, I will introduce my econometric model, and provide the
empirical analysis.
2. Literature Review
Although the impact of fiscal policy on women’s employment is well developed, the
impact of monetary policy on women’s employment is still a pretty new research
topic for macroeconomics and feminist economics. The main research papers on
this field are Abell (1991), Braunstein and Heintz (2006), Tachtamanova and
Sierminska (2009), and Seguino and Heintz (2010).
Abell (1991) indicates that distributional effects of monetary and fiscal policy on
unemployment are neither gender-neutral nor race-neutral. He focuses on 19741980, 1980-1987 periods in the US by using VAR methodology. White-male is the
most benefited group among them. According to his research, black females were
the other group that benefited significantly. Because of polarized job market, racial
and gender unemployment rates are not identical. White-males are employed in
primary sector while women, minorities, and youths are employed in secondary
sector.
As Elson (2007) also mentions, Braunstein and Heintz (2006) is the pioneering
study about the link between the monetary policies and gender equality in
employment. They mention the disproportioned effects of inflation reduction on
employment, and the difference of its impacts for women and men. Another
question they ask is that ‘how are monetary policy indicators connected to genderspecific employment effects’? They examine the relationship between inflation
3
reduction policies and men and women’s formal employment for 51 ‘inflation
reduction periods’ in 17 low and middle-income countries. They apply HodrikPrescott filter to the employment series in their estimation. Their results show that
contractionary monetary policy has negative disproportionate effects on women’s
employment. On the other hand, maintaining a competitive exchange rate can help
to reduce this impact. For non-contractionary monetary policy period, they do not
find a strong relationship between policy instrument and women’s formal
employment.
Tachtamanova and Sierminska (2009) investigate whether the policies for reducing
inflation cause disproportionate unemployment effects for women for nine OECD
countries between 1980 and 2004, using quarterly data. Those countries are
Canada, Finland, Italy, Japan, Norway, Spain, Switzerland, the US, and the UK so their
analysis focuses on developed countries. They use single equation regression and
vector autoregression analysis methodologies for economy wide, and sector
analysis. They conclude that the link between short-term interest rate and
employment is weak for the industrial countries that they researched, and it does
not vary by gender.
Seguino and Heintz (2010) focus on the impacts of contractionary monetary policy
on gender and race in the US from 1979-2008. They use federal funds rate, the
interest rate on overnight loans between banks, as the indicator of contractionary
monetary policy. In their paper, they examine ‘the impact of the federal funds rate
on the relative unemployment rates of different social groups’. Their econometrics
models are as follows for race and gender, respectively:
4
𝐵𝑊𝑀
𝑈𝑖𝑡𝐵𝑊𝑀 = 𝛽0 + 𝛽1 𝐹𝐹𝑅𝑡 + 𝛽2 𝐿𝐹𝑃𝑅𝑖𝑡
+ 𝛽3 𝑔𝑟𝑖𝑡 + 𝛽4 𝐵𝐿𝑆𝐻𝑖𝑡 + 𝛽5 𝐵𝐿𝑆𝐻𝑖𝑡2 + 𝜂𝑖 + 𝜀𝑖𝑡
𝐹𝑊𝑀
𝑈𝑖𝑡𝐹𝑊𝑀 = 𝛽0 + 𝛽1 𝐹𝐹𝑅𝑡 + 𝛽2 𝐿𝐹𝑃𝑅𝑖𝑡
+ 𝛽3 𝑔𝑟𝑖𝑡 + 𝛽4 𝐵𝐿𝑆𝐻𝑖𝑡 + 𝛽5 𝐵𝐿𝑆𝐻𝑖𝑡2 + 𝜂𝑖 + 𝜀𝑖𝑡
‘where U
BWM
represents the ratio of the total Black unemployment rate to the White
male unemployment rate; the subscripts i and t index states and years, respectively;
FFR
is federal funds rate (the nominal rate less the rate of inflation); LFPR
BWM
is the ratio of
Black to White male labor force participation rates; gr is the state-level growth of output
after the impact of national level growth dynamics have been removed; BLSH and BLSH
2
are the Black share of the population and Black share squared, respectively; 𝜂𝑖 is the
component of the disturbance term associated with state-specific effects; and 𝜀𝑖𝑡 is a
random error term’ (Seguino and Heintz 2010). In the second equation, superscript
FWM shows the ratio of female to male as difference from the first equation. According
to their results, monetary policy is neither gender nor race neutral. Unemployment,
as a result of contractionary monetary policy, affects black and white women more
negatively than white men. Moreover, black women have the burden more than
white women.
3. Contractionary Monetary Policy
In mid-1960s and 1970s, inflation was a big problem for Western economies
(Sargent, 1982: 41). Therefore after this period inflation reduction policies gained
importance for a lot of countries. The first country that adopted inflation targeting is
New Zealand in 1990. Idea behind these policies was that high inflation causes
damages for economy in the long run, and growth and employment can be
5
maintained through reducing inflation (Epstein, and Yeldan, 2008: 3). Therefore, the
primary object of monetary policy became maintaining price stability.
However, the other economic problems are also as significant as inflation. Akyuz
(2006:46 as cited in Epstein, and Yeldan, 2008: 5) mentions that main challenging of
macroeconomic policy is unemployment and financial instability, not inflation.
Furthermore, inflation might not be so dangerous unless it is too high. Inflation up
to 20% does not create problems in terms of growth, investment, foreign direct
investments or other real variables (Epstein, 2007: 7).
Dominance of these inflation reduction policies surpasses the other macroeconomic
targets. The results of reducing inflation policies can be against the other
macroeconomic policies. Reducing unemployment is one of the main targets among
these macroeconomic policies. Unemployment and informalization are the
opportunity costs of lower inflation (Jayadev, 2009: 75). The results of Ball’s (1997)
research shows that the reason of rising unemployment during 1980s is tight
monetary policy that aims to reduce inflation.
Inflation targeting policies has not been so successful about other sub-targets than
reducing the rate of inflation. The main objectives of these policies are:
• Reduce the rate of inflation
• Enhance the credibility of monetary policy
• Reduce the sacrifice ratio associated with contractionary monetary policy
• Help to attract foreign investment (Bernanke and others (1999), Mishkin and
Schmidt-Hebbel (2001) and Roger and Stone (2005) as cited in Epstein, 2007:4).
However, inflation targeting could neither enhance the credibility of monetary
6
policy nor reduce the sacrifice ratio (Bernanke and others, 1999, and Epstein, 2000,
as cited in Epstein, 2007:4). Tight monetary policy is used in order to keep inflation
in the low single digits, by using short-term interest rates as tool. Because of this
policy, many countries could not achieve the hoped for gains in employment rates or
economic growth (Epstein, 2007:1-7).
The contractionary monetary policy reduces inflation rate; however, it did not result
in expected gains in economic growth or employment (Epstein, Yeldan, 2008: 8).
According to Epstein and Yeldan’s (2008) research, the economic growth differs
among countries. They observe rise in unemployment for the post inflation
targeting period. In terms of balance of trade, there are both countries that
improved trade surplus and countries with huge deficit. After adoption of inflation
targeting, there was general trend towards appreciation for most countries caused
by the expansion of foreign capital inflows.
In the inflation target regime, monetary policy focuses on ‘setting the policy interest
rate’. Exchange rate is left to the global financial markets, and ‘floating/flexible’
exchange rate system is implemented. On the other hand, according to structuralist
tradition, it is important to have a stable and competitive real exchange rate because
it influences employment and economy in general (Epstein and Yeldan, 2008: 1516).
Macroeconomic policies are sometimes in favor of one group such as in terms of
class, gender, or race. In other words, macroeconomic policies have distributional
effects. Resent research show that monetary policy (especially contractionary
monetary policy) has also distributional effects on different groups. Workers are
7
influenced by negative impacts of higher unemployment more, and capitalists are
affected by higher inflation more negatively. Therefore workers are unemployment
averse while capitalists are inflation averse (Jayadav, 2009: 72). Moreover, as
mentioned
in
literature
review
part,
women
and
black
people
are
disproportionately influenced by contractionary policy in US (Seguino and Heintz,
2010).
To remove the inequality between women and men’s employment, we should
observe which policy is in favor of women, which policy is against women, or which
policy is gender neutral. If we can find a robust relationship between these policies
and women’s employment, it should be taken into account when preferring
macroeconomic policies.
4. Gender Differences in the Economy
Women face gender inequality in the economy, especially in terms of employment.
Elson (2007: 3-5) mentions the results of ILO 2004 report. The analysis shows that
for both 1993 and 2003, labor force participation rate and employment to
population rate is higher for men than women. However, female unemployment rate
is higher than male unemployment rate. Furthermore, in almost all regions
researched, occupations are segregated and there is an earning-gap in terms of
gender.
One of the explanations for the gender differences in the economy is the sexual
division of labor. The sexual division of labor both includes the pattern of work
allocation between women and men, and the social practices which classify some
8
works as suitable for women but unsuitable for men, or suitable for men but
unsuitable for women (Elson, 1995: 3).
We can observe the gender inequality in the economy as the segregation of women
in occupations. In sub-Saharan Africa and Southern Asia, women are primarily
employed in agricultural sector as ‘unpaid family workers’. For the rest of the
economies, women are usually concentrated in the service sector (UN 2000:114, as
cited in Elson, 2007: 5-6). Service sector is also segregated. Women work in
community, social and personal services, while men work in financial and business
services. In industry, women are primarily engage in food processing, textile, and
garment production (ILO 2004: 12 as cited in Elson, 2007: 6). Moreover, women are
concentrated in informal works more than men.
Tachtamanova and Sierminska (2009: 325-327) mention the reasons that can cause
the different effects of interest rate on female and male unemployment rate. The
causes that they discussed are employment and occupation segregation, gender
differences in labor market attachment, job tenure, and discrimination. They state
that empirical evidence shows that women and men works in different fields of
employment, and women has narrower range of occupation in the labor market.
They have different accesses to labor market attachment such as in the care
economy. Women are involved in full-time employment less then men, and they
have shorter tenure.
Employer’s preference is also influential in gender discrimination. Women are
usually considered as cheap labor force so especially some labor-intensive sectors
employ women to raise their profits.
9
5. Empirical Analysis
i. Data
This paper focuses on the impacts of contractionary monetary policy on women’s
employment for OECD countries between 1980-2014. In this analysis, short-term
interest rate shows the instrument of monetary policy, and I examine its effects on
the ratio of female unemployment rate to male unemployment rate. I use quarterly
data for 1980-2014 period as soon as data is available. Because data is not available
for whole period, I make unbalanced country-level panel data analysis.
My data sources are mainly OECD and ILOSTAT. For unemployment rates of women
and men, short term interest rates, GDP growth, and labor force participation rate of
women and men, I prefer OECD Labor Market Statistics, OECD Monthly Monetary
and Financial Statistics (MEI), OECD Quarterly National Accounts, ILOSTAT
databases, respectively.
I include all OECD countries except from Switzerland and Turkey. The reason why I
exclude Switzerland is that quarterly data for unemployment rate is only available
for the second quarters for Switzerland. I also exclude Turkey because short-term
interest rate data is not available for Turkey in OECD database. Therefore, I checked
Central Bank of Turkish Republic’s data but Central Bank data is not in the same
format with OECD quarterly data. Central Bank applies 1-week repo interest rate,
data of which is available from 2010 to 2014, as main policy interest rate. However,
‘the CB uses a policy mix comprising the interest rate of one-week repo auctions, the
interest rate corridor between overnight borrowing and lending rates and reserve
requirements’ (CBRT website). Therefore, a good reflection of short-term interest
10
rate requires an average interest rate of these three interest rates. However, the
data of these three types of interest rates are not available for same periods, and
they are not quarterly data. The data belongs to the day of decision. Therefore, I
leave Turkey for further research in which I am planning to find a method to take
the average interest rates.
Before starting my analysis I applied fisher-type unit root test based on augmented
Dickey-Fuller test for the variables. For all variables, I rejected the null hypothesis.
Therefore, there is no unit root.
ii. Econometric Model
As starting point, I will choose an econometric model similar to Seguino and Heintz
(2010). Because I will just focus on gender, I will not include black share of the
population or a similar indicator as an independent variable. My econometric model
is:
𝑟𝑎𝑡𝑖𝑜𝑢𝑛𝑖𝑡 = 𝛽0 + 𝛽1 shortterminterestrate𝑖𝑡 + 𝛽2 ratiolfpr𝑖,𝑡−1 + 𝛽3 𝑔𝑑𝑝𝑔𝑟𝑜𝑤𝑡ℎ𝑖,𝑡−1
+ 𝜂𝑖 + 𝜀𝑖𝑡
; where

𝑟𝑎𝑡𝑖𝑜𝑢𝑛: the ratio of the female unemployment rate to the male
unemployment rate

shortterminterestrate: ‘Short term rates are usually either the three month
interbank offer rate attaching to loans given and taken amongst banks for
any excess or shortage of liquidity over several months or the rate associated
with Treasury bills, Certificates of Deposit or comparable instruments, each
11
of three month maturity (OECD Monthly Monetary and Financial Statistics
(MEI))’.

ratiolfpr: the ratio of female to male labor force participation rate

𝑔𝑑𝑝𝑔𝑟𝑜𝑤𝑡ℎ: Growth rate compared to previous quarter, seasonally adjusted

𝜂𝑖 : the component of the disturbance term associated with country-specific
effects

𝜀𝑖𝑡 : random error term.
To observe whether there is a disproportionate impact, I will use the ratio of female
unemployment rate to male unemployment rate as dependent variable. My main
aim is to observe the impacts of policy interest rate on the ratio of female
unemployment rate to male unemployment rate. For beginning, I choose short-term
interest rate as a proxy of policy interest rate. The independent variable that I want
to observe its effects is short-term interest rate. I believe the ratio of labor force
participation rates has also effect on the ratio of unemployment rates. I also want to
use GDP growth rate as a control variable. I prefer to use fixed effect panel data
analysis to capture the country-specific effects, and 𝜂𝑖 shows the fixed effects in the
model. To avoid endogeneity problem, I lagged two variables: GDP growth and the
ratio of labor force participation rates.
iii. Results
I use country fixed effects regression. The regression is autocorrelation robust. The
results of the regression are reported in Table 1.
All results are statistically significant at the 1% level. According to the results of the
regression, the coefficient of short-term interest rate is statistically significant in the
12
1% level, and has a positive coefficient (0.0217). This result shows that the shortterm interest rate has negative impact on women’s employment. When central
banks prefer to apply contractionary monetary policy by increasing interest rates,
female unemployment rate increase more than male unemployment rate.
The other coefficients are also as expected. The coefficient of the ratio of labor force
participation rates (female/male) is positive (1.357), which shows that the more
women participate in the labor force (relative to men), the more women are
unemployed. Growth also has positive coefficient, which means with GDP growth,
women became more unemployed than men. GDP growth creates new job
opportunities more for men than women.
Table 1.
VARIABLES
shortterminterestrate
L.ratiolfpr
L.gdpgrowth
Constant
(1)
ratioun
0.0217***
(0.00171)
1.357***
(0.0491)
0.0146***
(0.00236)
-0.0704***
(0.0127)
Observations
1,859
Number of id_name
32
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
6. Concluding Remarks
The preliminary results show that there is a link between monetary policy and
women’s unemployment rate, and contractionary monetary policy has a
13
disproportionate negative effect on female employment for OECD countries. This
preliminary research about the links between economic policy instruments and
women employment requires further research. For further research, I am planning
to improve my econometric model. I believe an improved version of this
econometric model can reflect the relation between economic policy and women’s
employment better. There are several points that should be taken into account.
Firstly, in literature review part, we saw that there is a disproportionate impact of
monetary policy in terms of gender in developing countries; however,
Tachtamanova and Sierminska (2009) do not find such relationship for developed
countries. Therefore, another question that should be asked in the improved version
of econometric model is ‘are the impacts of contractionary monetary policies on
female unemployment rate different in developing and developed countries’? To
observe this difference, using a dummy variable for development level, or choosing
GDP per capita as an independent variable can be preferred. Gender-specific labor
market institutions can also differ between developing and developed countries.
Secondly, sectorial analysis can be added to the econometric model. The first step is
to observe in which sector women are working more intensely in OECD countries,
and the next step is to show that which sectors are affected more from slowdown of
employment resulting from contractionary monetary policy. Female intense sectors
could be more vulnerable to these macroeconomic shocks.
The period that examined includes the last global economic crises so the effects of
economic crisis can be influential in the unemployment rate. The global economic
crisis has severe impacts on some OECD countries so this also affects the result of
14
the empirical analysis of this paper. Therefore, this period can be examined into two
sub-periods: 1980-2008, 2008-2014.
European Central Bank applies one monetary policy for all Eurozone countries.
However, those countries have their own dynamics independent from European
Monetary Union. Therefore, examination of these countries requires special
attention. Again using a dummy variable for Eurozone countries can be solution for
better results.
In this paper, I research the impact of short-term interest rate on unemployment
rate, disaggregated in terms of gender. However, short-term interest rate is not the
only instrument of monetary policy. Moreover, there are other economic policies;
fiscal policy, and austerity policy, which also should be taken into account.
Therefore, the final question I am planning to clarify in the further research is ‘What
are the impacts of other economic policy instruments on the ratio of female
unemployment rate to male’s’?
15
Bibliography
Abell, J. D. (1991). Distributional effects of monetary and fiscal policy: impacts on
unemployment rates disaggregated by race and gender. American Journal of
Economics and Sociology, 50(3), 269-284.
Ball, L. M. (1997). Disinflation and the NAIRU. Reducing inflation: Motivation and
strategy (pp. 167-194). University of Chicago Press.
Braunstein, E., & Heintz, J. (2006). Gender bias and central bank policy: employment
and inflation reduction. International Review of Applied Economics, 22(2), 173-186.
Elson, D. (1995). Male bias in macro-economics: the case of structural adjustment.
Male bias in the development process, 164-90.
Elson, D. (2007). Macroeconomic Policy, Employment, Unemployment, and Gender
Equality. Towards Full and Decent Employment, 0-91.
Epstein, G. (2007). Central banks as agents of employment creation. Ocampo JA and
Jomo KS, Towards Full and Decent Employment. New York, London, Penang and
Hyderabad, Orient Longman Private Ltd, Zed Books, Third World Network and United
Nations, 92-122.
Epstein, G., & Yeldan, E. (2008). Inflation targeting, employment creation and
economic development: assessing the impacts and policy alternatives. International
Review of Applied Economics, 22(2), 131-144.
Heintz, J. (2006). Globalization, economic policy and employment: Poverty and gender
implications. Geneva: International Labour Office.
Heintz, J., & Seguino, S. (2011). Federal Reserve Policy and Inflation Dynamics in the
US: Race and Gender Inequalities in Unemployment Outcomes. Working paper,
Political Economy Research Institute and University of Vermont.
Jayadev, A. (2009). 4. Income, class and preferences towards anti-inflation and antiunemployment policies. Beyond Inflation Targeting: Assessing the Impacts and Policy
Alternatives, 71.
Sargent, T. J. (1982). The ends of four big inflations. In Inflation: Causes and effects
(pp. 41-98). University of Chicago Press.
Seguino, S. (2003). Why are women in the Caribbean so much more likely than men
to be unemployed?. Social and Economic Studies, 83-120.
Tachtamanova, Y. and E. Sierminska. (2009). “Gender, Monetary Policy, and
Employment: The Case of Nine OECD Countries.” Feminist Economics 15(3): 323-
16
353.
17